Brief reply to Allin:
If we're thinking about planning and socialism, we need to think about
shadow prices in a programming problem, and we wind up even further from
embodied labor coefficients, don't we?
I don't take the growing empirical evidence that market prices of
commodities are highly correlated with embodied labor coefficients both
in terms of level and change as addressing the labor theory of value
itself, but either a) the profit-rate equalizing theory of competition or
b) the actual structure of production (in terms of input-output tables).
That is, either labor is a dominant cost, so that variations in prices
are closely connected to variations in labor inputs, or profit rates are
not rapidly equalized according to this evidence.
Duncan